[postgis-devel] [mtnclimb at gmail.com: Re: Fwd: ST_UnaryUnion vs ST_Dump + ST_Union]
Sandro Santilli
strk at keybit.net
Mon Oct 3 00:18:21 PDT 2011
So it turns out it's a bug in GEOS.
See forwarded message from Martin Davis, and the ticket I've filed
for it: http://trac.osgeo.org/geos/ticket/482
This would call for a quick 3.3.2 release for sure.
Now if I only had time to make it really "quick" ...
--strk;
----- Forwarded message from Martin Davis <mtnclimb at gmail.com> -----
Date: Sun, 2 Oct 2011 13:44:17 -0700
From: Martin Davis <mtnclimb at gmail.com>
To: Sandro Santilli <strk at keybit.net>
Subject: Re: Fwd: [postgis-devel] ST_UnaryUnion vs ST_Dump + ST_Union
In-Reply-To: <20111002185015.GD2394 at xtp>
JTS works as expected - it returns
MULTILINESTRING ((0 0, 5 0), (5 0, 10 0, 5 -5, 5 0), (5 0, 5 5))
(Note this is different from the result that you suggested - but it seems
correct).
I will add this test for this to UnaryUnionTest.java. But you should also
look at TestUnaryUnion.xml - it has tests showing noding, as well as lots of
others.
Martin
On Sun, Oct 2, 2011 at 11:50 AM, Sandro Santilli <strk at keybit.net> wrote:
> On Sun, Oct 02, 2011 at 08:02:44PM +0200, Sandro Santilli wrote:
> > On Sun, Oct 02, 2011 at 08:35:44AM -0700, Martin Davis wrote:
>
> > UnaryUnionTest.java doesn't contain such a case.
> > I'll add it (single self-intersecting line) to the GEOS
> > port of that test, if you add it on the JTS side too
> > we'll keep 1:1 mapping.
>
> I confirm GEOS fails this new test:
>
> geom1_ = GEOSGeomFromWKT("LINESTRING(0 0,10 0,5 -5,5 5)");
> geom2_ = GEOSUnaryUnion(geom1_);
>
> ensure_equals(toWKT(geom2_), std::string(
> "MULTILINESTRING ((0 0,5 0),(5 0,10 0),(5 -5,5 0),(5 0,5 5))"
> ));
>
> Returns instead: "LINESTRING (0 0, 10 0, 5 -5, 5 5)"
>
> Does JTS work as expected ?
>
> --strk;
>
> () Free GIS & Flash consultant/developer
> /\ http://strk.keybit.net/services.html
>
----- End forwarded message -----
--
() Free GIS & Flash consultant/developer
/\ http://strk.keybit.net/services.html
More information about the postgis-devel
mailing list